Speaker
Description
Modern synchrotron facilities expose thousands of TANGO devices, requiring coordinated control of complex operations from simple queries to fast feedback loops. Elettra 2.0, the upcoming fourth-generation storage ring at Elettra-Sincrotrone Trieste, significantly increases this complexity with orders-of-magnitude higher brilliance, MHz-scale feedback rates, and greatly expanded network bandwidth, challenging traditional GUI- and script-based control paradigms.We present a Model Context Protocol (MCP) server that exposes the full TANGO device hierarchy as a typed, discoverable tool interface for Large Language Model agents, enabling device discovery, attribute access and command execution. The system serves as the TANGO integration layer within the Elettra 2.0 AI-assisted control architecture and is validated against the facility’s physics-based digital twin.
| Tags | Tango AI |
|---|